Alzheimer's Disease and Parkinson's 

Disease

Robert L. Nussbaum, M.D., and Christopher E. Ellis, Ph.D.

Alan E. Guttmacher, M.D., Editor, Francis S. Collins, M.D., Ph.D., Editor

The incidence of many common diseases is increased among the relatives of affected patients, but the pattern of inheritance rarely follows Mendel's laws. Instead, such common diseases are thought to result from a complex interaction among multiple predisposing genes and other factors, including environmental contributions and chance occurrences. Identifying the genetic contribution to such complex diseases is a major challenge for genomic medicine. However, as so clearly foreseen nearly 350 years ago by the English physiologist William Harvey,1 finding the genetic basis for rarer, mendelian forms of a disease may illuminate the etiologic process and pathogenesis of the more common, complex forms. This is illustrated in the progress made in understanding Alzheimer's disease and Parkinson's disease through the investigation of the rare, clearly inherited forms of these diseases. The molecular basis of neurodegenerative disorders was reviewed in theJournal in 1999.2

ALZHEIMER'S DISEASE

The most common neurodegenerative disease, Alzheimer's disease constitutes about two thirds of cases of dementia overall (ranging in various studies from 42 to 81 percent of all dementias), with vascular causes and other neurodegenerative diseases such as Pick's disease and diffuse Lewy-body dementia making up the majority of the remaining cases.3,4

Alzheimer's disease is a progressive neurologic disease that results in the irreversible loss of neurons, particularly in the cortex and hippocampus.5 The clinical hallmarks are progressive impairment in memory, judgment, decision making, orientation to physical surroundings, and language. Diagnosis is based on neurologic examination and the exclusion of other causes of dementia; a definitive diagnosis can be made only at autopsy. The pathological hallmarks are neuronal loss, extracellular senile plaques containing the peptide β amyloid, and neurofibrillary tangles; the latter are composed of a hyperphosphorylated form of the microtubular protein tau.6Amyloid in senile plaques is the product of cleavage of a much larger protein, the β-amyloid precursor protein, by a series of proteases, the α-, β-, and γ-secretases.7 The γ-secretase, in particular, appears to be responsible for generating one particular β-amyloid peptide — Aβ42 — that is 42 amino acids in length and has pathogenetic importance, because it can form insoluble toxic fibrils and accumulates in the senile plaques isolated from the brains of patients with Alzheimer's disease.8,9

Measures of the prevalence of Alzheimer's disease differ depending on the diagnostic criteria used, the age of the population surveyed, and other factors, including geography and ethnicity.10,11Excluding persons with clinically questionable dementia, Alzheimer's disease has a prevalence of approximately 1 percent among those 65 to 69 years of age and increases with age to 40 to 50 percent among persons 95 years of age and over10 (Figure 1FIGURE 1Prevalence of Alzheimer's Disease as a Function of Age in Men and Women.). Although the mean age at the onset of dementia is approximately 80 years,3 early-onset disease, defined arbitrarily and variously as the illness occurring before the age of 60 to 65 years, can occur but is rare. In one community-based study in France, the prevalence of early-onset disease (defined by an onset before the age of 61 years) was 41 per 100,000; thus, early-onset cases make up about 6 to 7 percent of all cases of Alzheimer's disease.12 About 7 percent of early-onset cases are familial, with an autosomal dominant pattern of inheritance and high penetrance.12 Thus, familial forms of early-onset Alzheimer's disease, inherited in an autosomal dominant manner, are rare; however, their importance extends far beyond their frequency, because they have allowed researchers to identify some of the critical pathogenetic pathways of the disease.

Missense mutations that alter a single amino acid and therefore gene function have been identified in three genes in families with early-onset autosomal dominant Alzheimer's disease. Family linkage studies and DNA sequencing identified mutations responsible for early-onset autosomal dominant forms of the disease in the gene encoding β-amyloid precursor protein itself on chromosome 21 (Figure 2FIGURE 2Altered Amino Acid Residues in a Segment of the β-Amyloid Precursor Protein Adjacent to Its Transmembrane Domain Resulting from Missense Mutations and Causing Early-Onset Familial Alzheimer's Disease.), as well as in two genes with similarity to each other, presenilin 1 (PSEN1) on chromosome 14 and presenilin 2 (PSEN2) on chromosome 1.PSEN1 mutations are more common than PSEN2 mutations. In a study of French families, for example, half of patients with familial, early-onset Alzheimer's disease that was inherited as an autosomal dominant trait had mutations in PSEN1, whereas approximately 16 percent of families had mutations in the β-amyloid precursor protein (βAPP) gene itself.12 PSEN2mutations were not found, and the genes responsible for the remaining 30 percent or so of cases were unknown.

The presenilin and βAPP mutations found in patients with familial early-onset Alzheimer's disease appear to result in the increased production of Aβ42, which is probably the primary neurotoxic species involved in the pathogenesis of the disease7,13 (Figure 3FIGURE 3The Normal Processing of β-Amyloid Precursor Protein as Well as the Effect on Processing of Alterations in the Protein Resulting from Missense Mutations Associated with Early-Onset Familial Alzheimer's Disease.). In these forms of Alzheimer's disease, mutations in βAPP itself or in the presenilins can shift the cleavage site to favor the γ-secretase site14 and, in particular, to favor increased production of the toxic Aβ42 peptide over the shorter, less toxic Aβ40peptide. Presenilin 1 may in fact be the γ-secretase itself or a necessary cofactor in γ-secretase activity.15 The toxic peptide is increased in the serum of patients with various βAPP, PSEN1, and PSEN2 mutations causing early-onset Alzheimer's disease.16 Cultured cells transfected in order to express the normal β-amyloid precursor protein generally process approximately 10 percent of the protein into the toxic Aβ42 peptide. Expression of various mutant βAPP or PSEN1 genes associated with early-onset familial Alzheimer's disease can result in an increase in the production of Aβ42 by a factor of up to 10.17-19 The identification of mutations in βAPP and the presenilins in early-onset familial Alzheimer's disease not only suggests a common mechanism through which mutations in these genes may exert their deleterious effects (i.e., increased production or decreased clearance of Aβ42 and formation of a protein aggregate, the amyloid plaque) but also provides evidence of a direct role of the Aβ42peptide and presenilins in the pathogenesis of the disease.20 In contrast, mutations in the tau gene, which encodes a protein contained within another neuropathologic structure in Alzheimer's disease, the neurofibrillary tangle, have not been identified in families with hereditary Alzheimer's disease, although they are seen in another, rarer neurodegenerative disorder, frontotemporal degeneration with parkinsonism21,22 (Figure 3).

As important as the rare familial early-onset forms of Alzheimer's disease have been for understanding the pathogenesis of the disease, the majority of patients of any age have sporadic (nonfamilial) disease in which no mutation in the βAPP or presenilin genes has been identified. However, another genetic risk factor, variants of APOE, the gene that encodes apolipoprotein E, a constituent of the low-density lipoprotein particle, has been associated with Alzheimer's disease.23Three variants of the gene and the protein are found in human populations and result from changes in single amino acids in apolipoprotein E (referred to as the APOE ε2, ε3, and ε4 alleles). Carrying one APOE ε4 allele nearly doubles the lifetime risk of Alzheimer's disease (from 15 percent to 29 percent), whereas not carrying an APOE ε4 allele cuts the risk by 40 percent.24 Initially, survival curves analyzing the effect of the APOE ε4 allele on the occurrence of Alzheimer's disease suggested that 70 to 90 percent of persons without this allele were disease-free at the age of 80 years, whereas 30 to 60 percent of those with one APOE ε4 allele and only 10 percent of homozygous persons surviving to the age of 80 were disease-free.23 A more recent study also provided evidence that APOE ε4 has a role in Alzheimer's disease, but the effect was less marked, with the rate of disease-free survival as high as 70 percent in homozygous persons.25

Although the magnitude of the effect of the APOE ε4 allele differs among studies, there appears to be a dose effect, in that disease-free survival was lower in homozygous persons than in heterozygous persons. This observation has led to the conclusion that the primary effect of theAPOE ε4 allele is to shift the age at onset an average of approximately 5 to 10 years earlier in the presence of one allele and 10 to 20 years earlier in the presence of two alleles in persons with an underlying susceptibility to Alzheimer's disease.26 The molecular mechanisms by which the various APOE alleles alter the age at onset and, therefore, the lifetime risk of Alzheimer's disease are unknown. A number of associations of the disease with variants of genes other than APOEhave also been reported but remain to be confirmed and are the subject of ongoing research.27

Because of the relative rarity of βAPP, PSEN1, and PSEN2 mutations in patients with late-onset Alzheimer's disease, we believe that molecular testing for mutations in these genes should be restricted to those with an elevated probability of having such mutations — that is, persons with early-onset disease or a family history of the disease. At-risk, symptomatic relatives of persons with documented mutations in βAPP or one of the presenilins may also request testing for the purposes of family, financial, or personal planning. Testing of a presymptomatic person should be undertaken with extreme care and only after extensive pretest counseling, so that the person requesting the test is aware of the potential for severe psychological complications of testing positive for an incurable, devastating illness. There may also be serious ramifications in the area of employment and in obtaining life, long-term care, disability, or health insurance. Also important is that a positive test may indicate that other family members, who may not have participated in any counseling or consented to testing, will be identified as being at a substantially increased risk for early-onset Alzheimer's disease by virtue of their relationship to the person who tests positive.

The usefulness of testing for the APOE ε4 allele is also limited. Finding one or two APOE ε4 alleles in a symptomatic person with dementia certainly increases the likelihood that one is dealing with Alzheimer's disease and might be used as an adjunct to clinical diagnosis.28 On the other hand, since 50 percent of patients with autopsy-proved Alzheimer's disease did not carry an APOEε4 allele, its negative predictive value in a symptomatic person is very limited.24 APOE ε4 testing in asymptomatic persons has very poor positive and negative predictive values and should not be used.24

Insights derived from the identification of mutations in rare families with early-onset Alzheimer's disease are proving useful for identifying therapeutic targets and creating animal models for evaluating therapies.29 For example, β-secretase inhibitors have been developed and may prove useful in treating Alzheimer's disease by reducing Aβ42 production.30 Transgenic mice expressing mutant β-amyloid precursor protein have an age-dependent increase in the amount of Aβ42formation, increased plaque formation, and spatial memory deficits; they have, however, only a minimal loss of neurons.31 In addition, mice transgenic for both a βAPP and a PSEN1 mutation show accelerated deposition of Aβ42, as compared with mice expressing either transgene alone.32In transgenic mice with a mutant β-amyloid precursor protein, immunization with Aβ42 resulted in a decrease in plaque formation and an amelioration of memory loss.32-34 However, phase 2 clinical trials investigating immunization therapy with Aβ42 35 had to be suspended because of an increased risk of aseptic meningoencephalitis.35-37 In addition, other drugs such as statins, clioquinol, and certain nonsteroidal antiinflammatory drugs38 are being evaluated in mouse models of these rare, heritable forms of Alzheimer's disease.

PARKINSON'S DISEASE

Parkinson's disease is the second most common neurodegenerative disorder, after Alzheimer's disease. It is characterized clinically by parkinsonism (resting tremor, bradykinesia, rigidity, and postural instability)39 and pathologically by the loss of neurons in the substantia nigra and elsewhere in association with the presence of ubiquinated protein deposits in the cytoplasm of neurons (Lewy bodies)40,41 and thread-like proteinaceous inclusions within neurites (Lewy neurites). Parkinson's disease has a prevalence of approximately 0.5 to 1 percent among persons 65 to 69 years of age, rising to 1 to 3 percent among persons 80 years of age and older.42 The diagnosis is made clinically, although other disorders with prominent symptoms and signs of parkinsonism, such as postencephalitic, drug-induced, and arteriosclerotic parkinsonism, may be confused with Parkinson's disease until the diagnosis is confirmed at autopsy.43

A genetic component in Parkinson's disease was long thought to be unlikely, because most patients had sporadic disease and initial studies of twins showed equally low rates of concordance in monozygotic and dizygotic twins.44 The view that genetics was involved in some forms of Parkinson's disease was strengthened, however, by the observation that monozygotic twins with an onset of disease before the age of 50 years do have a very high rate of concordance — much higher than that of dizygotic twins with early-onset disease.44,45 Furthermore, regardless of the age at onset, the apparent rate of concordance among monozygotic twins can be significantly increased if abnormal striatial dopaminergic uptake in the asymptomatic twin of a discordant pair, as revealed by positron-emission tomography with fluorodopa F18, is used as a sign of presymptomatic Parkinson's disease.46,47 An increased risk of Parkinson's disease was also seen among the first-degree relatives of patients,48,49 particularly when the results of positron-emission tomography of asymptomatic relatives were taken into account,50 providing further evidence of the existence of a genetic component to the disease.

However, the real advance occurred when a small number of families with early-onset, Lewy-body–positive autosomal dominant Parkinson's disease were identified.51 Investigation of these families, of Mediterranean and German origin, led to the identification of two missense mutations (Ala53Thr and Ala30Pro) in the gene encoding α-synuclein, a small presynaptic protein of unknown function.52,53 Although mutations in α-synuclein have proved to be extremely rare in patients with Parkinson's disease, they did provide the first clue that this protein could be involved in the molecular chain of events leading to the disease. The importance of α-synuclein was greatly enhanced by the discovery that the Lewy bodies and Lewy neurites found in Parkinson's disease in general contain aggregates of α-synuclein54,55 (Figure 4FIGURE 4Immunohistochemical Analysis of Sections from the Substantia Nigra of a Patient with Sporadic Parkinson's Disease, Indicating the Involvement of α-Synuclein in the Formation of Lewy Bodies and Lewy Neurites.). Molecules of α-synuclein protein are prone to form into oligomers in vitro; proteins carrying the missense mutations Ala53Thr and Ala30Pro seem to be even more prone to do so.56

Although the study of families with early-onset Parkinson's disease proves that abnormal α-synuclein can cause the disease, it is still unclear whether fibrils of aggregated α-synuclein, as seen in Lewy bodies and Lewy neurites, have a critical causative role in the more common forms of Parkinson's disease or are simply a marker for the underlying pathogenetic process. Lewy bodies positive for α-synuclein are found not only in various subnuclei of the substantia nigra, the locus ceruleus, and other brain-stem and thalamic nuclei of patients with Parkinson's disease, but also in a more diffuse distribution, including the cortex in some patients with Parkinson's disease as well as in patients with dementia of the diffuse Lewy-body type.57,58 Aggregated α-synuclein in glia is also a feature of multiple-system atrophy,59 leading to the coining of a new nosologic term, “synucleinopathy,” to refer to the class of neurodegenerative diseases associated with aggregated α-synuclein.60

Autosomal recessive juvenile parkinsonism is another genetic neurologic syndrome that has provided important insights into Parkinson's disease. Autosomal recessive juvenile parkinsonism is a relatively rare syndrome that shares many of the features of parkinsonism, including responsiveness to levodopa and loss of nigrostriatal and locus ceruleus neurons, but it has a very early onset (before the age of 40 years), a slow clinical course extending over many decades, and no Lewy bodies or Lewy neurites at autopsy.61 Genetic mapping of the syndrome to 6q25–27 led to the identification of mutations responsible for autosomal recessive juvenile parkinsonism in a gene encoding a protein termed parkin.62 Parkin is expressed primarily in the nervous system and is one member of a family of proteins known as E3 ubiquitin ligases, which attach short ubiquitin peptide chains to proteins, a process termed ubiquination, thereby tagging them for degradation through the proteosomal degradation pathway.

Autosomal recessive juvenile parkinsonism results from a loss of function of both copies of the parkin gene,63-65 leading to autosomal recessive inheritance, as opposed to the missense mutations that alter α-synuclein and cause a dominantly inherited disorder. More recently, however, the spectrum of disease known to be caused by parkin mutations has broadened, with apparently sporadic Parkinson's disease occurring in adulthood, as late as in the fifth and sixth decades of life, in association with parkin gene mutations.66 There have even been a few patients with apparently classic sporadic Parkinson's disease with an onset in adulthood who appear to have only one mutant parkin allele, although an exhaustive demonstration that the other allele is truly normal and not harboring an unusual mutation outside the coding sequence and its immediate environs is still lacking. Precisely what role parkin mutations have in the majority of cases of Parkinson's disease and whether the heterozygous state (which is far more common in the population than is homozygosity for two mutant alleles) represents an important risk factor remain to be established.

Recent evidence suggests that ubiquination by parkin may be important in the normal turnover of α-synuclein.67 Also of interest is the finding in one family of a few members with Parkinson's disease who had a deleterious missense mutation in the gene encoding a neuron-specific C-terminal ubiquitin hydrolase, another gene involved in ubiquitin metabolism.68 The obvious inference from these disparate pieces of data is that aggregation of abnormal proteins, dysfunctional ubiquitin-mediated degradation machinery, or both may be important steps in the pathogenesis of Parkinson's disease.

In addition to the α-synuclein, parkin, and ubiquitin C-hydrolase genes, at least five other loci have been proposed for autosomal dominant69-71 and autosomal recessive72-74 Parkinson's disease (Table 1TABLE 1Mutations in Single Genes That Lead to Parkinson's Disease.). Genetic analysis of the more common, sporadic forms of Parkinson's disease suggests that there is a component of heritability in the forms that are not clearly inherited as autosomal dominant or recessive traits.75-78 For example, certain alleles at a complex DNA-repeat polymorphic locus approximately 10 kilobase pairs upstream of the α-synuclein gene have been shown to be associated with sporadic Parkinson's disease in some populations, but not in others.79-82 Positive identification of the genes at these loci is likely to provide additional genes and proteins that can be studied for their roles in the pathogenesis of the disease.

Because of the extreme rarity of α-synuclein mutations, genetic testing for these mutations should be performed only on a research basis when a strong family history of autosomal dominant Parkinson's disease is present. Homozygous parkin mutations are found in the nearly half of patients presenting with apparent Parkinson's disease in childhood and adolescence and perhaps 5 percent of young adults with Parkinson's disease.64 There is little evidence supporting a role for mutations in the parkin gene in typical late-onset Parkinson's disease, and neither α-synuclein nor parkin gene testing is currently available as a routine clinical service.

CONCLUSIONS

The common neurodegenerative diseases are predominantly idiopathic disorders of unknown pathogenesis. As the examples of Alzheimer's disease and Parkinson's disease demonstrate, however, the genetic mapping and gene-isolation tools created by the Human Genome Project over the past decade have greatly accelerated the rate of identification of genes involved in the rare inherited forms of these diseases and are now being used to determine the genetic contributions to the more common, multifactorial forms of these diseases. The emergence of a consensus hypothesis — aggregates of Aβ42 and α-synuclein are neurotoxic in Alzheimer's disease and Parkinson's disease, respectively — may explain the pathogenesis not only of the inherited forms of these diseases but also of the idiopathic variety. Such insights into causation and pathogenesis are helping to identify new treatment targets for these debilitating d